VOGONS

Search

Common searches

The curious case of the artifacting Geforce 2 MX

Topic actions

First post, by botond87

Posted on 
User metadata
Rank Newbie
Rank
Newbie

Hi there,

I recently grabbed a first revision Geforce 2 MX for my FIC-VB601 & Pentium II 400 rig, booted it up, got into Win98…and almost immediately started seeing major artifacts, that got so bad, that I couldn’t identify anything on the screen about 2 minutes in.

After powering off and booting up again, I entered the BIOS and waited to see if the artifacts appear again - sure enough, after about 2 minutes in, the screen started collapsing again with artifacts getting worse every passing second.

I thought to myself okay, since there is no heatsink on this first revision model, I’ll just strap a small heatsink to it with a bit of thermal paste (the chip was getting pretty hot to the touch after about a minute) and test it again. The heatsink also got pretty warm, but the artifacts still destroyed the screen after 2 minutes.

At this point I thought okay, the seller sold me a fried card (although claiming that it’s in working condition), so out of rage and a bit of curiosity, I planned to install some drivers, and downclock the card, to see if that would bring back some life into it. So I switched the GF2MX to my GF4, and installed the version 30.82 reference drivers. Then I swapped back to the GF2MX and just managed to point in the device manager to the driver folder, before the screen fell apart.

Now comes the curious part: to my surprise, after rebooting the rig, the artifacts were gone. The card was recognized by the driver successfully, everything seemed to be in order. I waited on the desktop for a few minutes to see if the artifacts would come back - nope.

Then I got a bit cocky, and fired up Giants to make the card squeal, played for 30 minutes straight - no issues whatsoever.

I’m still planning to do a proper stress test, but this behaviour still puzzles me. I’ve never seen a card that presented artifacts - even in the BIOS! - and then just started working normally after a fresh driver install.

If I wasn’t desperate enough to get a driver installed, I would’ve easially just dropped the card into my junk pile, and called it a day.

At this point, this is just an interesting story that I wanted so share with you all, but I still don’t get it, how the card manages to work.

TL;DR: card with massive artifacts suddenly works after clean driver install.

Did anyone have a similar experience, does anyone know how this is possible?

Thank you for reading, cheers! 🙂

Reply 1 of 7, by drosse1meyer

Posted on 
User metadata
Rank Member
Rank
Member

i had strange problems with a geforce - artifacting and such, which went away with downclocking.

ultimately a full recap fixed all the issues at normal clock speed in both 2d and GL modes.

P1: Packard Bell - 233 MMX, Voodoo1, 64 MB, ALS100+
P2-V2: Dell Dimension - 400 Mhz, Voodoo2, 256 MB
P!!! Custom: 1 Ghz, GeForce2 Pro/64MB, 384 MB

Reply 2 of 7, by cyclone3d

Posted on 
User metadata
Rank l33t++
Rank
l33t++

I wonder if the edge connector just needs to be cleaned.

Oxidation can mess with signaling as well as proper voltage getting to a card.

Could also be that the capacitors are on their way out and they started working after being power cycled a few times.

Yamaha modified setupds and drivers
Yamaha XG repository
YMF7x4 Guide
Aopen AW744L II SB-LINK

Reply 3 of 7, by botond87

Posted on 
User metadata
Rank Newbie
Rank
Newbie

The connections (as well as the card) looked pretty clean, but I guess it’s possible that there was some layer of oxidation, which I’ve rubbed of by taking it out and putting it back 6-7 times 😄

I’ll also keep an eye on the caps, physically they look fine to me, but yeah, they are 23 years old.

Thanks for the tips guys!

Reply 4 of 7, by Tetrium

Posted on 
User metadata
Rank l33t++
Rank
l33t++
botond87 wrote on 2023-03-26, 13:59:

The connections (as well as the card) looked pretty clean, but I guess it’s possible that there was some layer of oxidation, which I’ve rubbed of by taking it out and putting it back 6-7 times 😄

I’ll also keep an eye on the caps, physically they look fine to me, but yeah, they are 23 years old.

Thanks for the tips guys!

Connectors may look clean, but still be covered by a lot of filth (the filth will show afterwards on the item you used to clean the contacts with, you'd be surprised how much can come off even though the contacts look clean at first).
Cleaning the contacts is a good, cheap and easy way to re-try seemingly defective cards (and also memory modules).

Whats missing in your collections?
My retro rigs (old topic)
Interesting Vogons threads (links to Vogonswiki)
Report spammers here!

Reply 5 of 7, by danieljm

Posted on 
User metadata
Rank Member
Rank
Member

Have you tried sitting in the BIOS again for a couple minutes to see if the problem still happens? The drivers won't be loaded, so you can see if the problem still exists or if it magically fixed itself.

If the problem is still there, I wonder if there's some corruption in the VBIOS that has clocks set too high, and when the driver loads it brings them back in line. I have no idea if this is a thing that could happen, I'm just trying to think outside the box. 😀

Reply 6 of 7, by botond87

Posted on 
User metadata
Rank Newbie
Rank
Newbie

In case anyone was wondering - the solution was replacing the caps. I’ve just got around to ordering some replacement caps for the 3 largest ones on the card - a 470 uF + 10 uF electrolytic and a 470uF tantalum cap.

After replacing them, the issue was completely gone - the card is now artifact-free!

So drosse1meyer was right about the caps failing, apparently until they didn’t get to “operating tempearature” or didn’t get enough load, they were not up to spec.

Reply 7 of 7, by drosse1meyer

Posted on 
User metadata
Rank Member
Rank
Member
botond87 wrote on 2023-07-26, 08:26:

In case anyone was wondering - the solution was replacing the caps. I’ve just got around to ordering some replacement caps for the 3 largest ones on the card - a 470 uF + 10 uF electrolytic and a 470uF tantalum cap.

After replacing them, the issue was completely gone - the card is now artifact-free!

So drosse1meyer was right about the caps failing, apparently until they didn’t get to “operating tempearature” or didn’t get enough load, they were not up to spec.

glad you got it working. another vintage part saved from the dustbin of history
👍

P1: Packard Bell - 233 MMX, Voodoo1, 64 MB, ALS100+
P2-V2: Dell Dimension - 400 Mhz, Voodoo2, 256 MB
P!!! Custom: 1 Ghz, GeForce2 Pro/64MB, 384 MB

Go to top of page Go to top of page
Back to Video
Close window
Close window